We evaluated the association of BMI and multiple markers of male reproductive potential in a large group of men attending a fertility clinic and found that overweight and obesity were associated with abnormalities in serum levels of reproductive hormones and to a lesser extent with abnormalities in standard semen analysis and measures of sperm DNA integrity. However, overweight men had a slightly higher total progressive sperm count compared to normal weight men. Overall, these data suggest that despite marked changes in reproductive hormone levels with relatively small changes in body weight, only extreme levels of obesity may negatively influence male reproductive potential as assessed by semen quality and sperm DNA integrity.
We found strong inverse associations of BMI with serum levels of total testosterone, SHBG and inhibin B and a positive association with serum estradiol levels. These associations are well documented effects of excess body weight on these hormones. Excess adiposity leads to increased aromatization of androgens in the adipose tissue leading to higher circulating estradiol levels (42
). Hyperinsulinemia, secondary to obesity-related insulin resistance, decreases SHBG production in the liver (44
). Low testosterone levels are thought to be the result of decreased SHBG binding capacity (46
), direct action of leptin and other adipocyte-derived hormones on Leydig cells (47
) and, in morbidly obese men, impaired functioning of the hypothalamic-pituitary-testicular (HPT) axis (50
) possibly as a result of enhanced negative feedback on gonadotropin secretion by estradiol (52
). Overweight and obesity have been related to lower testosterone and SHBG levels and higher estradiol levels in multiple studies (26
) and body weight has been found to explain a greater proportion of the variability in testosterone levels than age and lifestyle practices (54
). Further, testosterone increases after weight loss in massively obese men (50
). Our findings regarding inhibin B levels are in agreement with four previous reports of the relationship between body weight and inhibin B in adult men (26
). Moreover, in a study among severely obese men who underwent gastroplasty, inhibin B levels increased after surgery among the men with the greatest amount of weight loss (on average 50 kg or 16.9 kg/m2
). The observed lower testosterone:LH ratio among the most obese men also suggests decreased Leydig cell function among these men and is consistent with a report of impaired LH-stimulated testosterone production among morbidly obese men (47
). The consistency of these findings across studies and the reversibility of this pattern following weight loss suggest a causal role of increased body weight on the hormonal pattern described above.
We also found inverse associations between BMI and gonadotropin levels which were more marked among men with abnormal semen analysis results. In men with an intact HPT axis lower levels of testosterone and inhibin B, as those observed with increasing levels of body weight, would be expected to result in higher levels of LH and FSH, respectively. Our findings suggest that excess body weight can lead to an impairment of the feedback regulation of the HPT axis, particularly among men who eventually develop semen quality abnormalities. Several studies have reported no relation between excess body weight and gonadotropin levels (26
). Yet, our interpretation is in agreement with reports of decreased LH pulse amplitude (46
), decreased total LH secretion over a 12 hour period (51
) and increased LH levels following weight loss among massively obese men (50
). Similarly, total FSH secretion over a 24 hour period decreases with increasing body weight in men (52
) and plasma FSH levels increase after massive weight loss (56
). Previous investigations have not examined whether the effect of body weight on gonadotropin secretion may differ according to other personal characteristics, as suggested by our results. This possibility should be further evaluated in other studies.
We did not observe statistically significant differences in sperm concentration, sperm morphology or sperm motility across levels of BMI. Only ejaculate volume was significantly lower in overweight and obese men relative to normal weight men. In addition, total sperm count (ejaculate volume × sperm concentration) was significantly lower in the group of most obese men (BMI ≥ 35 kg/m2
); a difference that could be explained to some extent by our results for ejaculate volume. Furthermore, we found that overweight men had a slightly higher total progressive sperm count than normal weight men. This could represent a chance finding given that the past two studies reporting on the relation between body weight or abdominal adiposity and progressive motility have found that this parameter decreases with increasing adiposity (60
). Others have reported that increased adiposity is related to decreased fertility (29
) and negatively affects nearly every semen analysis parameter including concentration (26
), ejaculate volume (61
), total sperm count (26
), motile count (61
) and progressive motility (60
). The most consistent positive finding across studies has been lower sperm concentration among overweight and obese men compared to normal weight men. This has been reported by five previous studies (26
) while another three (27
) did not find this association. When our results are included with these past studies, almost just as many studies have reported a null association between overweight and obesity and sperm concentration as have been studies reporting lower sperm concentration with increased body weight. Null findings on other parameters have been more consistent. Our null findings regarding the potential role of BMI on motility and morphology are in agreement with six of the seven past studies that have reported on motility (26
) and all the previous studies that have reported on morphology (26
). Similarly consistent, but in contrast with our results, have been reports of no association between adiposity and ejaculate volume (26
). Unfortunately, there does not appear to be any pattern in terms of study setting and size, participant personal characteristics or type of statistical analysis used that seem to differentiate between studies reporting deleterious effects of overweight and obesity on semen characteristics from those that do not. An additional complication is that some studies have reported their results in a way that is not easy to interpret and does not necessarily imply compromised spermatogenesis. For example, Kort and colleagues reported an inverse association between BMI and total normal-motile spermatozoa count (volume × concentration × %motility × %normal morphology) (30
). Because results for the individual parameters were not reported it is not possible to know which were affected and complicates interpretation as a significant difference in any one parameter could explain the association with this composite outcome. Clearly more studies are needed in this area to clarify the role of body weight on semen quality.
We found that obese men, but not overweight men, had a greater number of sperm cells with high DNA damage as assessed with the comet assay. However, there was no relation between BMI and the other three standard measures of sperm DNA integrity in this assay. Only one study has previously reported on the relationship between BMI and sperm DNA integrity. Using the SCSA assay to assess chromatin integrity, Kort and collaborators found that overweight and obese men had a significantly higher percentage of sperm with DNA damage when compared to normal weight men (30
). It is important to note, however, that SCSA and the comet assay measure different aspects of sperm DNA integrity (SCSA measures susceptibility of sperm chromatin to DNA denaturation while the comet assay measures the extent of sperm DNA fragmentation in individual sperm). Because of this, and since BMI was unrelated to the other three measures of sperm DNA integrity in our study, the apparent consistency between these two studies should be viewed with caution.
Strengths of our study include the direct assessment of anthropometric measures and our ability to account for multiple potential confounders neither of which has been the case in some previous studies. A salient limitation is the fact that only a single measure of hormone levels and semen analysis are available. Nevertheless, despite the circadian, pulsatile and circannual variation in the levels of specific reproductive hormones, a single blood sample can provide an adequate measure of testosterone over a year in adult men (68
) and the between-person variation in testosterone, inhibin B, LH, FSH and SHBG serum levels is greater than their within-person variation over a 17-month period (69
) indicating that a single measure of these hormones may also adequately represent long-term levels. Further, obtaining multiple semen samples per subject in a population based study is not superior to obtaining a single semen sample (70
) and standard semen analysis parameters are stable over a 4 year period (72
In summary, we observed the well known relationships between body weight and reproductive hormone levels in a group of men attending an infertility clinic. Despite the significant differences in hormone levels, only obesity was associated with increased sperm DNA damage and only the most obese men (BMI ≥ 35 kg/m2) had a lower total sperm count when compared to normal weight men. These data suggest that differences in reproductive hormone levels due to increased body weight do not necessarily lead to impaired reproductive potential in men.